Papers
2
Total Citations
4
H-Index
2
About
Ruosong Ye is a researcher advancing the field of autonomous robot navigation through innovative applications of deep reinforcement learning (DRL). His work directly addresses critical limitations in how robots perceive and move through unknown environments. Ye’s primary research focus is on developing DRL frameworks that overcome the “local minimum” problem—a common failure mode where a robot becomes trapped in a suboptimal path. His most cited papers introduce two novel algorithms: the Adaptive Forward Simulation Time (AFST) and the Adaptive Execution Duration (AED), both operating within a Semi-Markov Model. These contributions are significant because they move beyond the standard, rigid approach of uniform command intervals. By allowing the robot to dynamically adjust how long it executes a command or how far it simulates a path, Ye’s methods enable more efficient and robust navigation in complex, cluttered spaces. Although his work is early-stage, with each of his top papers garnering 2 citations, the conceptual leap in adaptive timing represents a promising direction for making DRL-based navigation more practical and reliable for real-world robotic systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2